EP1933272B1 - Method of generating a contrast enhanced image using multiscale analysis - Google Patents

Description

FIELD OF THE INVENTION
• The present invention relates to a method for enhancing the contrast of an image that is represented by a digital signal.
BACKGROUND OF THE INVENTION
• Commonly images represented by a digital signal such as medical images are subjected to image processing during or prior to displaying or hard copy recording.
• The conversion of grey value pixels into values suitable for reproduction or displaying may comprise a multi-scale image processing method (also called multi-resolution image processing method) by means of which the contrast of the image is enhanced.
• According to such a multi-scale image processing method an image, represented by an array of pixel values, is processed by applying the following steps. First the original image is decomposed into a sequence of detail images at multiple scales and occasionally a residual image. Next, the pixel values of the detail images are modified by applying to these pixel values at least one conversion. Finally, a processed image is computed by applying a reconstruction algorithm to the residual image and the modified detail images. Such a method is illustrated in figures 2 and 3.
• There are limits for the behavior of the conversion functions. Grey value transitions in the image can be distorted to an extent that the appearance becomes unnatural if the conversion functions are excessively non-linear. The distortions are more pronounced in the vicinity of significant grey level transitions, which may result in overshoots at step edges and loss of homogeneity in regions of low variance facing strong step edges. The risk of creating artifacts becomes more significant for CT images since they have sharper grey level transitions, e.g. at the interface of soft tissue and contrast media. One has to be careful using the multi-scale techniques on CT images.
• In US 6 731 790 B1 a method is described in which artefacts are corrected in an image that is represented by a digital signal. The corrected artifacts originate from any kind of non-linear modification of the components of a multi-scale image representation.
• A gradient representation of the original image is generated and modified. A modification step is applied to gradient images.
• A reconstruction process is applied to the modified gradient representation.
• US 6 788 826 B1 teaches a multiscale decomposition of an image, a nonlinear modification of edge images and an inverting reconstruction algorithm. The weights used in the reconstruction algorithm are the normal weights of the decomposition algorithm.
• It is an object of the present invention to provide a method for enhancing the contrast of an image that is represented by a digital signal that overcomes the prior art inconveniences.
• More specifically it is an object of the present invention to provide a new multi-scale contrast enhancement algorithm which results in a contrast enhanced image while preserving the shape of the edge transitions.
SUMMARY OF THE INVENTION
• The above-mentioned aspects are realized by a method as defined in the appending claims.
• Specific and preferred embodiments of this invention are set out in the dependent claims.
• The invention provides that enhanced detail pixel values are created by enhancement of the center differences.
• The method of the present invention is different from the prior art multi-scale contrast enhancement for the following reason: the prior art algorithms amplify the detail pixel values by applying the conversion functions directly to the detail pixel values by means of look up tables or multiplicative amplification factors.
• The present invention is applicable to all the multi-scale detail representation methods from which the original image can be computed by applying the inverse transform.
  The invention is applicable to all the multi-scale decomposition methods wherein the detail pixel are equivalent to the sum of translation difference images or can be computed as a center difference image.
• According to the present invention enhanced detail pixel values are directly computed by combining the enhanced center differences at at least one scale. (illustrated in figures 4, 6 and 8 with corresponding enhancement steps being illustrated in figures 5, 7 and 9 respectively)
• The overall characteristics of the present invention are shown in figure 1.
• The present invention is generally implemented as a computer program product adapted to carry out the method of any of the claims when run on a computer and is stored on a computer readable medium.
• Further advantages and implementations of the embodiments of the present invention will be explained in the following description and will be illustrated by the drawings.
• In the context of the present invention specific terms are defined as follows:
• Multi-scale decomposition mechanism:
  • A multi-scale (or multi-resolution) decomposition of an image is a process that computes detail images of said image at multiple scales of a grey value image. A multi-scale decomposition mechanism generally involves filter banks for computing the detail images. Well-known techniques are for example: the Laplacian pyramid, the Burt pyramid, the Laplacian stack, the wavelet decomposition, QMF filter banks.
• Approximation image:
  • An approximation image is a grey value image that represents the original grey value image at the same or a larger scale, or at the same or a lower resolution. An approximation image at a specific scale is equivalent to the original grey value image in which all details at that scale have been omitted (Mallat S.G.,"A Theory for Multiresolution Signal Decomposition: The Wavelet Representation", IEEE Trans. On Pattern Analysis and Machine Intelligence, vol. 11, no. 7, July 1989).
• Detail image:
  • A detail image is defined as the difference of information between an approximation image at a certain scale and an approximation image at a smaller scale.
• Conversion operator:
  • A conversion operator is operator which generates the pixel-wise modification of the detail pixel values as an intermediate step to create a contrast enhanced version of the grey value image. Such an operator has for example been described in European patent EP 527 525 . The modification is defined by a conversion function and can e.g. be implemented as a look up table or as a multiplicative amplification.
• Translation difference image:
  • The translation difference images at a scale s are a measurement of elementary contrast in each pixel of an approximation image at scale s. They can be computed by taking the difference of the approximation image at that scale s and a translated version. Other computations for elementary contrast are possible, e.g. the ratio of pixel with a neighboring pixel can be used in case the processing steps are preceded by an exponential transform and followed by a logarithmic transform.
• Center difference image:
  • A center difference image is computed by applying a combining operator (for example the summation) to translation difference images.
  • The combining operator can be a linear or non-linear function of corresponding pixel values in the translation difference images.
BRIEF DESCRIPTION OF THE DRAWINGS
• Fig. 1 schematically illustrates the steps of the method of the present invention,
• Fig. 2 illustrates a multi-resolution image processing method according to the prior art,
• Fig. 3 illustrates the image enhancement step of the multi-resolution image processing method illustrated in figure 2,
• Fig. 4, 6 and 8 illustrate different implementations of the multi-resolution image processing method according to the present invention,
• Fig. 5 illustrates the image enhancement step of the multi-resolution image processing method illustrated in figure 4,
• Fig. 7 illustrates the image enhancement step of the multi-resolution image processing method illustrated in figure 6,
• Fig. 9 illustrates the image enhancement step of the multi-resolution image processing method illustrated in figure 8,
• Fig. 10 illustrates the number of translation difference images and the translations needed in the method of the present invention,
• Fig. 11 is a legend pertaining to the symbols used in the above figures.
DETAILED DESCRIPTION OF THE INVENTION
• This contrast enhancement algorithm is applicable to all multi-scale detail representation methods from which the original image can be computed by applying the inverse transformation.
• It is applicable to the reversible multi-scale detail representation that can be computed as a weighted sum of translation difference images.
  The weighing factors and the translation offsets of the translation difference images can be deducted from the multi-scale decomposition in such a way that the resulting weighted sum of the translation difference images is identical to the detail pixel values.
• For these multi-scale detail representations the contrast can be enhanced by applying the conversion operator to the center differences before the weighted sum is computed.
• To compute the weighted sum of translation difference images, the approximation image at the same scale (or resolution level) or the approximation images at the smaller scales (or finer resolution levels) can be used.
• State-of-the-art multi-scale contrast enhancement algorithms decompose an image into a multi-scale representation comprising detail images representing detail at multiple scales and a residual image.
  Some of the important multi-scale decompositions are the wavelet decomposition, the Laplacian-of-Gaussians (or LoG decomposition), the Difference-of-Gaussians (or DoG) decomposition and the Burt pyramid.
• The wavelet decomposition is computed by applying a cascade of high-pass and low-pass filters followed by a subsampling step.
• The high-pass filter extracts the detail information out of an approximation image at a specific scale.
• In the Burt pyramid decomposition the detail information is extracted out of an approximation image at scale k by subtracting the upsampled version of the approximation image at scale k+ 1.
• In a state of the art methods as the one disclosed in EP 527 525 a contrast enhanced version of the image is created by conversion of the pixel values in the detail images followed by multi-scale reconstruction.
• All above implementations of multiscale decomposition have a common property. Each pixel value in the detail images can be computed out of an approximation image by combining the pixel values in a moving neighborhood.
  In the above cases the combining function is a weighted sum.
• For the wavelet decomposition the pixel values in the detail image at scale k are computed as: $${\mathit{d}}_{\mathit{k}\mathit{+}\mathit{1}}\mathit{=}\mathit{\downarrow }\left({\mathit{h}}_{\mathit{d}}\mathit{*}{\mathit{g}}_{\mathit{k}}\right)$$

  

  $$g_{\mathit{k}\mathit{+}\mathit{1}}\mathit{=}\mathit{\downarrow }\left(l_{\mathit{d}}\mathit{*}{\mathit{g}}_{\mathit{k}}\right)$$

  

  with h_d a high-pass filter, l_d a low-pass filter, * the convolution operator and ↓ the subsampling operator (i.e. leaving out every second row and column).
• For the wavelet reconstruction the enhanced approximation image at scale k is computed as: $${\mathit{h}}_{\mathit{k}}\mathit{=}{\mathit{l}}_{\mathit{r}}\mathit{*}\left(\mathit{\uparrow }{\mathit{h}}_{\mathit{k}\mathit{+}\mathit{1}}\right)\mathit{+}{\mathit{h}}_{\mathit{r}}\mathit{*}\left(\mathit{\uparrow }\mathit{f}\left({\mathit{d}}_{\mathit{k}\mathit{+}\mathit{1}}\right)\right)$$

  

  with h_r a high-pass filter, l_r a low-pass filter and ↗ the upsampling operator (i.e. inserting pixels with value 0 in between any two rows and columns).
• For the Burt decomposition the pixel values in the detail image at scale k are computed as: $${\mathit{d}}_{\mathit{k}}\mathit{=}{\mathit{g}}_{\mathit{k}}\mathit{-}\mathit{4}\mathit{g}\mathit{*}\left(\mathit{\uparrow }{\mathit{g}}_{\mathit{k}\mathit{+}\mathit{1}}\right)$$

  

  or $${\mathit{d}}_{\mathit{k}}\mathit{=}{\mathit{g}}_{\mathit{k}}\mathit{-}\mathit{4}\mathit{g}\mathit{*}\left(\mathit{\uparrow }\left(\mathit{\downarrow }\left(\mathit{g}\mathit{*}{\mathit{g}}_{\mathit{k}}\right)\right)\right)$$

  

  or $${\mathit{d}}_{\mathit{k}}\mathit{=}\left(\mathit{1}\mathit{-}\mathit{4}\mathit{g}\mathit{*}\left(\mathit{\uparrow }\left(\mathit{\downarrow }\mathit{g}\right)\right)\right)\mathit{*}{\mathit{g}}_{\mathit{k}}$$

  

  with g a Gaussian low-pass filter and 1 the identity operator.
• For the Burt reconstruction the enhanced approximation image at scale k is computed as: $${\mathit{h}}_{\mathit{k}}\mathit{=}\mathit{4}\mathit{g}\mathit{*}\left(\mathit{\uparrow }{\mathit{h}}_{\mathit{k}\mathit{+}\mathit{1}}\right)\mathit{+}\mathit{f}\left({\mathit{d}}_{\mathit{k}}\right)$$

  

  with f( x) the conversion operator.
The multi-scale detail pixel values as weighted sums
• Suppose that in the Burt multi-scale decomposition a 5x5 Gaussian filter is used with coefficients w_k,l with k = -2,...2 and 1 = -2,...,2, the subsampling operator removes every second row and column and the upsampling operator inserts pixels with value 0 in between any two rows and columns.
  The pixel at position i,j in the approximation image g_k+1 is computed as: $$g_{k+1}\left(ij\right)={\displaystyle \sum _{s=-2\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="imgf0006.png,imgf0008.png"\; state="\{\{state\}\}">t</figure-callout>}}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(2i+s,2j+t\right)$$

  

  The pixel at position i,j in the upsampled image u_k is computed as: $$u_k\left(ij\right)=\{{\displaystyle \sum _{s=-2\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="imgf0006.png,imgf0008.png"\; state="\{\{state\}\}">t</figure-callout>}}^2}{\displaystyle \sum _{t=-2}^2}\begin{array}{ll}{w}_{s,t}{g}_k\left(2i+s,2j+t\right)& \mathit{if\; i\; and\; j\; are\; even}\\ 0& \mathit{othervise}\end{array}$$

  
• The pixel at position i,j in the upsampled, smoothed image gu_k is computed as: $${\mathit{gu}}_k\left(ij\right)=\{\begin{array}{c}{\displaystyle \sum _{m=\left\{-2,0,2\right\}}}{\displaystyle \sum _{n=\left\{-2,0,2\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; and\; j\; are\; even}\\ {\displaystyle \sum _{m=\left\{-1,1\right\}}}{\displaystyle \sum _{n=\left\{-2,0,2\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; is\; odd\; and\; j\; is\; even}\\ {\displaystyle \sum _{m=\left\{-2,0,2\right\}}}{\displaystyle \sum _{n=\left\{-1,1\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; is\; even\; and\; j\; is\; odd}\\ {\displaystyle \sum _{m=\left\{-1,1\right\}}}{\displaystyle \sum _{n=\left\{-1,1\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; and\; j\; are\; odd}\end{array}$$

  
• Finally, the pixel at position i,j in the detail image d_k is computed as: $$d_k\left(ij\right)=\{\begin{array}{c}{g}_k\left(ij\right)-4{\displaystyle \sum _{m=\left\{-2,0,2\right\}}}{\displaystyle \sum _{n=\left\{-2,0,2\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; and\; j\; are\; even}\\ g_k\left(ij\right)-4{\displaystyle \sum _{m=\left\{-1,1\right\}}}{\displaystyle \sum _{n=\left\{-2,0,2\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; is\; odd\; and\; j\; is\; even}\\ g_k\left(ij\right)4{\displaystyle \sum _{m=\left\{-2,0,2\right\}}}{\displaystyle \sum _{n=\left\{-1,1\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; is\; even\; and\; j\; is\; odd}\\ g_k\left(ij\right)-4{\displaystyle \sum _{m=\left\{-1,1\right\}}}{\displaystyle \sum _{n=\left\{-1,1\right\}}}{w}_{m,n}{\displaystyle \sum _{s=-2}^2}{\displaystyle \sum _{t=-2}^2}{w}_{s,t}{g}_k\left(i+s+m,j+t+n\right)\mathit{if\; i\; and\; j\; are\; odd}\end{array}$$

  
• Generally, the pixel at position i,j in the detail image d_k can be computed as a weighted sum of pixels in the approximation image at the same or smaller scale k, k-1, k-2,... : $${\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\mathit{=}{\mathit{g}}_l\left(\mathit{ri}\mathit{rj}\right)\mathit{-}{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right)$$

  

  with 1∈{0, ..,k} and r = subsampling_factor^(1-k)
• Because $${\displaystyle \sum _m}{\displaystyle \sum _n}{v}_{m,n}=1$$

  

  the pixel at position i,j in the detail image d_k can be computed as: $${\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\mathit{=}{\mathit{g}}_l\left(\mathit{ri}\mathit{rj}\right)\mathit{-}{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right)$$

  

  $${\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\mathit{=}{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{rj}\right)\mathit{-}{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right)$$

  

  $${\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\mathit{=}{c}_k\left(ij\right)={\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}\mathit{=}\left({\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{rj}\right)-{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right)\right)$$

  
• The term g_l (ri,rj) - g_l (ri+m,rj+n) is called a translation difference.
  It expresses the difference in pixel value between a central pixel and a neighboring pixel in an approximation image. It is a measure of local contrast.
  The weighted sum of the translation differences is called a centre difference c_k (i,j).
• In a similar way it can be proven that the detail images in other multi-scale decomposition methods can also be represented as a combination of translation difference images.
The conversion operation
• In state-of-the-art methods like the one disclosed in EP 527 525 contrast enhancement is obtained by applying a conversion operator f(x) to the detail image d_k or, equivalently: $$f\left({\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\right)\mathit{=}f(g_l\left(\mathit{ri}\mathit{rj}\right)-{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}{\mathit{g}}_{\mathit{l}}\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right))$$

  
• An example of such a conversion operator is the sigmoid function. Another example of such conversion operator is the contrast enhancement function like the one disclosed in EP 525 527 . The shape of the conversion operator depends on the specific requirements of the enhancement which is intended to amplify the low-value detail pixel more than the high-value detail pixels.
• The conversion step may cause deformations of the shape of the edge transitions in the reconstructed, contrast enhanced image. The reason is the non-linearity of the conversion function.
• Generally, the following applies to non-linear functions: $$f\left(x+y\right)\ne f\left(x\right)+f\left(y\right)$$

  

  or $$f\left({\displaystyle \sum _i}{x}_i\right)\ne {\displaystyle \sum _i}f\left(x_i\right)$$

  
• State-of-the-art algorithms first compute the pixel values in the detail image d_k as weighted sums and apply the conversion step afterwards.
• By rewriting the pixel values in the detail image d_k as a weighted sum of translation differences, it is possible to apply the conversion step before the summation instead of afterwards.
• Contrast enhancement is now obtained by applying the conversion step to the translation differences: $$f\left({\mathit{d}}_{\mathit{k}}\left(\mathit{i}\mathit{j}\right)\right)\mathit{=}{\displaystyle \mathit{\sum }_{\mathit{m}}}{\displaystyle \mathit{\sum }_{\mathit{n}}}{\mathit{v}}_{\mathit{m}\mathit{,}\mathit{n}}\mathit{f}\left({\mathit{g}}_{\mathit{l}}\left(\mathit{ri},\mathit{rj}\right)-g_l\left(\mathit{ri}\mathit{+}\mathit{m}\mathit{,}\mathit{rj}\mathit{+}\mathit{n}\right)\right))$$

  
• In this way the shape of the edge transitions is better preserved in the contrast enhanced, reconstructed image.
• If for every scale k the detail image at that scale is computed out of the full resolution image g₀ , and enhancement is applied to the center differences, then the shapes of the edge transitions are best preserved after reconstruction.
• Different implementations of the present invention are illustrates in figures 4, 6 and 8 with the corresponding enhancement steps being illustrated in figures 5, 7 and 9.
• First the translation difference images are computed. This is done by subtracting a translated version of the approximation image g₁ from the approximation image g₁ itself.
  The number of translation difference images and the translations needed, are defined by the spatial extent of the cascade of filters used in the multi-scale decomposition to compute a detail image at scale k out of an approximation image at scale 1 element of {0, ...k}.
  For example, in case of a 5x5 Gaussian filter in the Burt decomposition and given the fact that the approximation image at the same scale as the detail image is used, 9x9 = 81 translation difference images are needed.
  Figure 10 schematically shows the 1D situation.
• For each pixel position i,j in the detail image the weights w_{m, n} are computed such that the weighted sum of the translation differences matches exactly the pixel value as would be generated by straightforward application of the Burt decomposition as described higher.
• Contrast enhancement is obtained by non-linear conversion of the translation difference images.
• At least one enhanced center difference image is computed as the weighted sum of the contrast enhanced translation difference images, using the weights w_m,n.
• The reconstructed, contrast enhanced version of the image is computed by applying the multi-scale reconstruction to the enhanced center difference images.
• The condition for the weights w_m,n as formulated above is required for the multi-scale decomposition/reconstruction algorithm disclosed above to be reversible in case the conversion step is excluded.

Claims (9)

1. A method of enhancing the contrast of an image that is represented by a digital signal wherein

   - at least one approximation image is created at one or multiple scales, by applying a multi-scale decomposition algorithm to said image, an approximation image at a scale representing the grey values of said image in which all details at that scale referred to as corresponding detail image, have been omitted,

   - translation difference images are created by pixel-wise subtracting the values of an approximation image at scale s and the values of translated versions of said approximation image,

   - the values of said translation difference images are nonlinearly modified,

   - at least one enhanced center difference image at a specific scale is computed by combining said modified translation difference images at that scale or a smaller scale with weights w_i,j wherein said weights w_i,j are selected in accordance with said image decomposition algorithm such that, if said enhancement by non-linear modification were not performed, an equivalent of said multi-scale decomposition of said image would be obtained such that said at least one center difference image would equal the detail image at said specific scale,

   - an enhanced image is computed by applying the image reconstruction algorithm which reverses said image decomposition algorithm to the image decomposition that results after said detail image at said specific scale has been replaced by said at least one enhanced center difference image.
2. A method according to claim 1 wherein a translation difference image at a specific scale is computed out of an approximation image at the same scale.
3. A method according to claim 1 modified in that all of said translation difference images are computed out of the original image.
4. A method according to claim 1 wherein a translation difference image at a scale k is computed out of an approximation image at scale m, wherein m represents a scale between scale 1 and scale k-1.
5. A method according to claim 1 or 4 wherein the center difference images are identical to the multi-scale detail images.
6. A method according to any of the preceding claims wherein said image is a mammographic image.
7. A method according to any of the preceding claims wherein said image is a CT image.
8. A computer program product adapted to carry out the steps of the method of any of the preceding claims when run on a computer.
9. A computer readable medium comprising computer executable program code adapted to carry out the steps of the method of any of the claims 1-7 when executed on a computer.

Images